FE-674: Planning sync after V3.1 shipped#125
Conversation
V3.1 implementation closed end-to-end on PR #124. This sync reflects that across the canonical planning docs. PLAN.md: - Move §Active item 1 (V3.0 cascade) to §Recently Completed — already shipped via FE-674 PRs #115-#118. - Promote continuous-workspace from §Next to §Active (item 1). - Move V3.1 from §Horizon to §Recently Completed with full ship note; remove the duplicate §Horizon entry. - Promote V4a side-chat-persistence into §Next (item 2), pointing at FE-675 V4a half. - Add V4b item-versioning + branched-exploration to §Horizon, gated on FE-701 + V4a. - Trim §Recently Completed to 3 most recent entries. - Update Track B dependency tree: V3.0 + V3.1 both marked completed; V4a stacked under V3.1; V4b stacked under semantic-changeset ledger. PLAN_HISTORY.md: - Append 2026-05-11 Sync Archive section with the five entries rotated out of §Recently Completed (V3.1 slice 4 reconciliation classifier, FE-674 planning sync, FE-698 prompt/context scenario substrate, V2 plumbing, V1.1 Explore vertical slice, graph-view structured-list). SIDE_CHAT.md §13: - Mark V3.0 and V3.1 as shipped with the actual PR references. - Split V4 row into V4a (next, FE-675) and V4b (horizon, FE-675/FE-701). - Rewrite §5.3 V3.1 prose to reflect the shipped classifier vocabulary, lifecycle, per-class actions, and bulk affordances. Note that A88 outer-loop walkthrough is still the qualitative validation step. CARDS.md: not on main; intentionally left absent (frontier item is now continuous-workspace, not yet card-scoped). No SPEC.md changes — V3.1 outcomes are already covered by D139 + I114 shipped earlier in the stack; no new D### or I### required by the V3.1 implementation as delivered. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
PR SummaryLow Risk Overview Reorders the live roadmap: promotes continuous workspace to Reviewed by Cursor Bugbot for commit 71dc802. Bugbot is set up for automated code reviews on this repo. Configure here. |
🤖 Augment PR SummarySummary: Documentation-only sync to close the V3.x side-chat arc and align the planning ledger with what shipped in FE-674. Changes:
Technical Notes: No 🤖 Was this summary useful? React with 👍 or 👎 |
![augmentcode[bot]](https://avatars.githubusercontent.com/in/1027498?s=60&v=4){kind=small}
| **V3.0 grouping is mechanical, not agent-classified.** Needs are grouped in the overlay by `kind` (`supersedes` first, then `needs_confirmation`) and within each kind by relation type. There is no auto-confirm / auto-edit / substantive ML classification in V3.0. | ||
|
|
||
| **V3.1 — agent-grouped resolution.** Once the reconciliation agent ships (MULTI_CHAT.md Phase 3), it reclassifies open needs into: | ||
| **V3.1 — agent-grouped resolution (shipped, FE-674 PR #124).** The reconciliation classifier (server `POST /reconciliation-needs/run-agent`) walks every open need through the lifecycle `null → queued → classifying → classified | failed` (I114) and persists exactly one of: |
There was a problem hiding this comment.
POST /reconciliation-needs/run-agent here doesn’t match the spec-scoped path referenced elsewhere in this PR (e.g. POST /api/specifications/:id/reconciliation-needs/run-agent); consider aligning the docs to one canonical endpoint to avoid confusion.
Severity: low
Other Locations
docs/archive/PLAN_HISTORY.md:175
🤖 Was this useful? React with 👍 or 👎, or 🚀 if it prevented an incident/outage.
|
|
||
| ## 2026-05-11 Sync Archive | ||
|
|
||
| Archived out of `memory/PLAN.md` after side-chat V3.1 closed end-to-end (PR #124) and the live ledger narrowed to V3.0 + V3.1 + the multi-chat substrate. Five older entries rotated out. |
There was a problem hiding this comment.
This paragraph says “Five older entries rotated out,” but the section below lists six archived bullets; consider reconciling the count so the archive note stays accurate.
Severity: low
🤖 Was this useful? React with 👍 or 👎, or 🚀 if it prevented an incident/outage.
kostandinang
changed the title
1 participant
What
Planning sync after V3.1 shipped. Updates
memory/PLAN.md,docs/archive/PLAN_HISTORY.md, anddocs/design/SIDE_CHAT.md§13 to match production. No code.Should land on
mainfirst so downstream FE-674 + FE-675 stacks rebase onto a shared baseline.Changes
No SPEC.md changes — V3.1 outcomes are covered by existing tracked rows.
Watch
Outer-loop walkthrough on a dense spec (does agent grouping help legibility?) hasn't happened yet. Captured as a Watch on the V3.1 entry.
Test plan